Classification of Gas Discharge Tube’s Electromagnetic Pulse Response Based on Kmeans Method

Gas discharge tube is a kind of common electromagnetic pulse protection device, which has strong non-linearity. Machine learning method is one of the methods of numerical modeling of gas discharge tube. In the process of machine learning modeling, the modeling precision on some data needs to be improved. Therefore, it is necessary to identify the different stages of the gas discharge tube first, and then to model it in segments to improve the overall modeling accuracy. K-means is a clustering method in machine learning, which can accomplish data clustering automatically when the type of data is unknown. This automatic recognition method first estimates the optimal number of categories of all modeling data. Secondly, k-means method is used to identify data based on the optimal category number. Experiments show that the method can automatically identify the different response stages of the gas discharge tube. The result of clustering can be used for the segmental modeling of the gas discharge tube in the future.